enlightenment/src/modules/pager/e_mod_main.c

2874 lines
78 KiB
C
Raw Normal View History

#include "e.h"
#include "e_mod_main.h"
/* gadcon requirements */
2006-05-31 09:19:11 -07:00
static E_Gadcon_Client *_gc_init(E_Gadcon *gc, const char *name, const char *id, const char *style);
static void _gc_shutdown(E_Gadcon_Client *gcc);
static void _gc_orient(E_Gadcon_Client *gcc, E_Gadcon_Orient orient __UNUSED__);
static const char *_gc_label(E_Gadcon_Client_Class *client_class __UNUSED__);
static Evas_Object *_gc_icon(E_Gadcon_Client_Class *client_class __UNUSED__, Evas *evas);
static const char *_gc_id_new(E_Gadcon_Client_Class *client_class __UNUSED__);
/* and actually define the gadcon class that this module provides (just 1) */
static const E_Gadcon_Client_Class _gadcon_class =
{
GADCON_CLIENT_CLASS_VERSION,
"pager",
{
List of changes: 1. Gadcon sities: -- 'E_Gadcon_Site' enum to list all available sities, -- added 'is_site' callback for distinguish gadcon sities, -- 'e_gadcon_site_is_*' helpers written to use in gadget modules, -- gadcon config dialogs modified to show only gadgets that appropriate to selected gadcon site, -- all gadget modules updated for using new callback. When callback is not provided in module, then gadget can be used in any gadcon site. 2. Filemanager: -- added error icon to unexisting path dialog, -- added callback 'func.abort' to 'E_Fm2_Op_Registry_Entry' to specify abort operation sequence (if not provided, then operation is not cancelable), -- added parameter to 'e_fm2_op_registry_entry_add' to specify abort method on construct 'E_Fm2_Op_Registry_Entry' or make non-cancelable operation if this function is not provided, -- added 'e_fm2_operation_abort' call to cancel efm2 operation, -- added method '_e_fm2_operation_abort_internal' that represent efm2 default handler of operation cancelling, -- 'done' and 'total' fields of 'E_Fm2_Op_Registry_Entry' changed to off_t type, also all message handlers modified to work with it (to show progress for large files correctly), -- improved file operation progress processing, -- added cancel button to operation progress gadget in filemanager window, -- added delete file glyph to operation progress gadget, -- improved status message for efm2 operation progress. 3. Filemanager slave: -- fixed critical bug in message handling via stdin/stdout, -- fixed critical bug in removing task from list, -- improved 'E_FM_OP_ABORT' handling, -- fix rollback counting on moving files. 4. Fileman_opinfo module: -- a new theme writed: status string and gadget for view operation status in summary and detail modes, -- summary mode: operation glyph, state message, progressbar, abort button, -- detail mode: all for summary + from filename, to filename, -- summary/detailed button works as trigger, -- added source of module theme and used images. 4. Pathbar module: -- non-critical fix, just for safety. Also patch contains several minor updates to serve main task. SVN revision: 41224
2009-06-29 07:38:03 -07:00
_gc_init, _gc_shutdown, _gc_orient, _gc_label, _gc_icon, _gc_id_new, NULL,
e_gadcon_site_is_not_toolbar
},
E_GADCON_CLIENT_STYLE_INSET
};
/* actual module specifics */
typedef struct _Instance Instance;
typedef struct _Pager Pager;
typedef struct _Pager_Desk Pager_Desk;
typedef struct _Pager_Win Pager_Win;
typedef struct _Pager_Popup Pager_Popup;
struct _Instance
{
E_Gadcon_Client *gcc;
Evas_Object *o_pager; /* table */
Pager *pager;
};
struct _Pager
{
Instance *inst;
E_Drop_Handler *drop_handler;
Pager_Popup *popup;
Evas_Object *o_table;
E_Zone *zone;
int xnum, ynum;
Eina_List *desks;
unsigned char dragging : 1;
unsigned char just_dragged : 1;
Evas_Coord dnd_x, dnd_y;
Pager_Desk *active_drop_pd;
};
struct _Pager_Desk
{
Pager *pager;
E_Desk *desk;
Eina_List *wins;
Evas_Object *o_desk;
Evas_Object *o_layout;
int xpos, ypos, urgent;
int current : 1;
struct
{
Pager *from_pager;
unsigned char in_pager : 1;
unsigned char start : 1;
int x, y, dx, dy, button;
} drag;
};
struct _Pager_Win
{
E_Border *border;
Pager_Desk *desk;
Evas_Object *o_window;
Evas_Object *o_icon;
unsigned char skip_winlist : 1;
struct
{
Pager *from_pager;
unsigned char start : 1;
unsigned char in_pager : 1;
unsigned char no_place : 1;
unsigned char desktop : 1;
int x, y, dx, dy, button;
} drag;
};
struct _Pager_Popup
{
E_Popup *popup;
Pager *pager;
Evas_Object *o_bg;
Ecore_Timer *timer;
unsigned char urgent : 1;
};
static void _pager_cb_obj_moveresize(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info __UNUSED__);
static void _button_cb_mouse_down(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info);
static void _menu_cb_post(void *data __UNUSED__, E_Menu *m __UNUSED__);
static void _pager_inst_cb_menu_configure(void *data __UNUSED__, E_Menu *m __UNUSED__, E_Menu_Item *mi __UNUSED__);
static void _pager_inst_cb_menu_virtual_desktops_dialog(void *data, E_Menu *m __UNUSED__, E_Menu_Item *mi __UNUSED__);
static void _pager_instance_drop_zone_recalc(Instance *inst);
static Eina_Bool _pager_cb_event_border_resize(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_move(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_add(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_remove(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_iconify(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_uniconify(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_stick(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_unstick(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_desk_set(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_stack(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_icon_change(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_urgent_change(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_focus_in(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_focus_out(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_border_property(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_zone_desk_count_set(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_desk_show(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_desk_name_change(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_cb_event_container_resize(void *data __UNUSED__, int type __UNUSED__, void *event);
static void _pager_window_cb_mouse_down(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info);
static void _pager_window_cb_mouse_up(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info);
static void _pager_window_cb_mouse_move(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info);
static void *_pager_window_cb_drag_convert(E_Drag *drag, const char *type);
static void _pager_window_cb_drag_finished(E_Drag *drag, int dropped);
static void _pager_drop_cb_enter(void *data, const char *type __UNUSED__, void *event_info);
static void _pager_drop_cb_move(void *data, const char *type __UNUSED__, void *event_info);
static void _pager_drop_cb_leave(void *data, const char *type __UNUSED__, void *event_info __UNUSED__);
static void _pager_drop_cb_drop(void *data, const char *type, void *event_info);
static void _pager_inst_cb_scroll(void *data);
static void _pager_update_drop_position(Pager *p, Evas_Coord x, Evas_Coord y);
static void _pager_desk_cb_mouse_down(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info);
static void _pager_desk_cb_mouse_up(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info);
static void _pager_desk_cb_mouse_move(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info);
static void _pager_desk_cb_drag_finished(E_Drag *drag, int dropped);
static void _pager_desk_cb_mouse_wheel(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info);
static Eina_Bool _pager_popup_cb_timeout(void *data);
static Pager *_pager_new(Evas *evas, E_Zone *zone);
static void _pager_free(Pager *p);
static void _pager_fill(Pager *p);
static void _pager_empty(Pager *p);
static Pager_Desk *_pager_desk_new(Pager *p, E_Desk *desk, int xpos, int ypos);
static void _pager_desk_free(Pager_Desk *pd);
static Pager_Desk *_pager_desk_at_coord(Pager *p, Evas_Coord x, Evas_Coord y);
static void _pager_desk_select(Pager_Desk *pd);
static Pager_Desk *_pager_desk_find(Pager *p, E_Desk *desk);
static void _pager_desk_switch(Pager_Desk *pd1, Pager_Desk *pd2);
static Pager_Win *_pager_window_new(Pager_Desk *pd, E_Border *border);
static void _pager_window_free(Pager_Win *pw);
static void _pager_window_move(Pager_Win *pw);
static Pager_Win *_pager_window_find(Pager *p, E_Border *border);
static Pager_Win *_pager_desk_window_find(Pager_Desk *pd, E_Border *border);
static Pager_Popup *_pager_popup_new(E_Zone *zone, int keyaction);
static void _pager_popup_free(Pager_Popup *pp);
static Pager_Popup *_pager_popup_find(E_Zone *zone);
static E_Config_Dialog *_pager_config_dialog(E_Container *con, const char *params);
/* functions for pager popup on key actions */
static int _pager_popup_show(void);
static void _pager_popup_hide(int switch_desk);
static Eina_Bool _pager_popup_cb_mouse_down(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_popup_cb_mouse_up(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_popup_cb_mouse_move(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_popup_cb_mouse_wheel(void *data __UNUSED__, int type __UNUSED__, void *event);
static void _pager_popup_desk_switch(int x, int y);
static void _pager_popup_modifiers_set(int mod);
static Eina_Bool _pager_popup_cb_key_down(void *data __UNUSED__, int type __UNUSED__, void *event);
static Eina_Bool _pager_popup_cb_key_up(void *data __UNUSED__, int type __UNUSED__, void *event);
static void _pager_popup_cb_action_show(E_Object *obj __UNUSED__, const char *params __UNUSED__, Ecore_Event_Key *ev __UNUSED__);
static void _pager_popup_cb_action_switch(E_Object *obj __UNUSED__, const char *params, Ecore_Event_Key *ev);
/* variables for pager popup on key actions */
static E_Action *act_popup_show = NULL;
static E_Action *act_popup_switch = NULL;
static Ecore_X_Window input_window = 0;
static Eina_List *handlers = NULL;
static Pager_Popup *act_popup = NULL; /* active popup */
static int hold_count = 0;
static int hold_mod = 0;
static E_Desk *current_desk = NULL;
static E_Config_DD *conf_edd = NULL;
static Eina_List *pagers = NULL;
Config *pager_config = NULL;
static E_Gadcon_Client *
2006-05-31 09:19:11 -07:00
_gc_init(E_Gadcon *gc, const char *name, const char *id, const char *style)
{
Pager *p;
Evas_Object *o;
E_Gadcon_Client *gcc;
Instance *inst;
Evas_Coord x, y, w, h;
const char *drop[] =
{
"enlightenment/pager_win", "enlightenment/border",
"enlightenment/vdesktop"
};
inst = E_NEW(Instance, 1);
p = _pager_new(gc->evas, gc->zone);
p->inst = inst;
inst->pager = p;
o = p->o_table;
gcc = e_gadcon_client_new(gc, name, id, style, o);
gcc->data = inst;
inst->gcc = gcc;
inst->o_pager = o;
evas_object_geometry_get(o, &x, &y, &w, &h);
p->drop_handler =
e_drop_handler_add(E_OBJECT(inst->gcc), p,
_pager_drop_cb_enter, _pager_drop_cb_move,
_pager_drop_cb_leave, _pager_drop_cb_drop,
drop, 3, x, y, w, h);
evas_object_event_callback_add(o, EVAS_CALLBACK_MOVE,
_pager_cb_obj_moveresize, inst);
evas_object_event_callback_add(o, EVAS_CALLBACK_RESIZE,
_pager_cb_obj_moveresize, inst);
evas_object_event_callback_add(o, EVAS_CALLBACK_MOUSE_DOWN,
_button_cb_mouse_down, inst);
pager_config->instances = eina_list_append(pager_config->instances, inst);
return gcc;
}
static void
_gc_shutdown(E_Gadcon_Client *gcc)
{
Instance *inst;
inst = gcc->data;
pager_config->instances = eina_list_remove(pager_config->instances, inst);
e_drop_handler_del(inst->pager->drop_handler);
_pager_free(inst->pager);
E_FREE(inst);
}
static void
_gc_orient(E_Gadcon_Client *gcc, E_Gadcon_Orient orient __UNUSED__)
{
Instance *inst;
inst = gcc->data;
e_gadcon_client_aspect_set(gcc,
inst->pager->xnum * inst->pager->zone->w,
inst->pager->ynum * inst->pager->zone->h);
e_gadcon_client_min_size_set(gcc, 16, 16);
}
static const char *
_gc_label(E_Gadcon_Client_Class *client_class __UNUSED__)
{
return _("Pager");
}
static Evas_Object *
_gc_icon(E_Gadcon_Client_Class *client_class __UNUSED__, Evas *evas)
{
Evas_Object *o;
char buf[PATH_MAX];
o = edje_object_add(evas);
2007-03-24 16:12:23 -07:00
snprintf(buf, sizeof(buf), "%s/e-module-pager.edj",
e_module_dir_get(pager_config->module));
edje_object_file_set(o, buf, "icon");
return o;
}
static const char *
_gc_id_new(E_Gadcon_Client_Class *client_class)
{
char buf[PATH_MAX];
snprintf(buf, sizeof(buf), "%s.%d", client_class->name,
eina_list_count(pager_config->instances));
return strdup(buf);
}
static Pager *
_pager_new(Evas *evas, E_Zone *zone)
{
Pager *p;
p = E_NEW(Pager, 1);
p->inst = NULL;
p->popup = NULL;
p->o_table = e_table_add(evas);
e_table_homogenous_set(p->o_table, 1);
p->zone = zone;
_pager_fill(p);
pagers = eina_list_append(pagers, p);
return p;
}
static void
_pager_free(Pager *p)
{
_pager_empty(p);
evas_object_del(p->o_table);
pagers = eina_list_remove(pagers, p);
E_FREE(p);
}
static void
_pager_fill(Pager *p)
{
int x, y;
e_zone_desk_count_get(p->zone, &(p->xnum), &(p->ynum));
for (x = 0; x < p->xnum; x++)
{
for (y = 0; y < p->ynum; y++)
{
Pager_Desk *pd;
E_Desk *desk;
2005-10-11 10:22:39 -07:00
desk = e_desk_at_xy_get(p->zone, x, y);
pd = _pager_desk_new(p, desk, x, y);
if (pd)
{
p->desks = eina_list_append(p->desks, pd);
2005-09-05 15:14:06 -07:00
if (desk == e_desk_current_get(desk->zone))
_pager_desk_select(pd);
}
}
}
}
static void
_pager_empty(Pager *p)
{
while (p->desks)
{
_pager_desk_free(p->desks->data);
p->desks = eina_list_remove_list(p->desks, p->desks);
}
}
static Pager_Desk *
_pager_desk_new(Pager *p, E_Desk *desk, int xpos, int ypos)
{
Pager_Desk *pd;
Evas_Object *o, *evo;
E_Border_List *bl;
E_Border *bd;
pd = E_NEW(Pager_Desk, 1);
if (!pd) return NULL;
pd->xpos = xpos;
pd->ypos = ypos;
pd->urgent = 0;
pd->desk = desk;
e_object_ref(E_OBJECT(desk));
pd->pager = p;
o = edje_object_add(evas_object_evas_get(p->o_table));
pd->o_desk = o;
e_theme_edje_object_set(o, "base/theme/modules/pager",
2006-08-22 21:02:03 -07:00
"e/modules/pager/desk");
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
if (pager_config->show_desk_names)
edje_object_part_text_set(o, "e.text.label", desk->name);
else
edje_object_part_text_set(o, "e.text.label", "");
e_table_pack(p->o_table, o, xpos, ypos, 1, 1);
2005-10-11 10:22:39 -07:00
e_table_pack_options_set(o, 1, 1, 1, 1, 0.5, 0.5, 0, 0, -1, -1);
evo = (Evas_Object *)edje_object_part_object_get(o, "e.eventarea");
if (!evo) evo = o;
evas_object_event_callback_add(evo, EVAS_CALLBACK_MOUSE_DOWN,
_pager_desk_cb_mouse_down, pd);
evas_object_event_callback_add(evo, EVAS_CALLBACK_MOUSE_UP,
_pager_desk_cb_mouse_up, pd);
evas_object_event_callback_add(evo, EVAS_CALLBACK_MOUSE_MOVE,
_pager_desk_cb_mouse_move, pd);
evas_object_event_callback_add(evo, EVAS_CALLBACK_MOUSE_WHEEL,
_pager_desk_cb_mouse_wheel, pd);
evas_object_show(o);
2005-10-11 10:22:39 -07:00
o = e_layout_add(evas_object_evas_get(p->o_table));
pd->o_layout = o;
e_layout_virtual_size_set(o, desk->zone->w, desk->zone->h);
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
edje_object_part_swallow(pd->o_desk, "e.swallow.content", pd->o_layout);
evas_object_show(o);
bl = e_container_border_list_first(desk->zone->container);
while ((bd = e_container_border_list_next(bl)))
{
Pager_Win *pw;
2005-10-11 10:22:39 -07:00
if ((bd->new_client) || (bd->zone != desk->zone) ||
((bd->desk != desk) && (!bd->sticky)))
continue;
pw = _pager_window_new(pd, bd);
if (pw) pd->wins = eina_list_append(pd->wins, pw);
}
e_container_border_list_free(bl);
return pd;
}
static void
_pager_desk_free(Pager_Desk *pd)
{
Pager_Win *w;
evas_object_del(pd->o_desk);
evas_object_del(pd->o_layout);
EINA_LIST_FREE(pd->wins, w)
_pager_window_free(w);
e_object_unref(E_OBJECT(pd->desk));
E_FREE(pd);
}
static Pager_Desk *
_pager_desk_at_coord(Pager *p, Evas_Coord x, Evas_Coord y)
{
Eina_List *l;
Pager_Desk *pd;
EINA_LIST_FOREACH(p->desks, l, pd)
{
Evas_Coord dx, dy, dw, dh;
evas_object_geometry_get(pd->o_desk, &dx, &dy, &dw, &dh);
if (E_INSIDE(x, y, dx, dy, dw, dh)) return pd;
}
return NULL;
}
static void
_pager_desk_select(Pager_Desk *pd)
{
Eina_List *l;
Pager_Desk *pd2;
if (pd->current) return;
EINA_LIST_FOREACH(pd->pager->desks, l, pd2)
{
if (pd == pd2)
{
pd2->current = 1;
evas_object_raise(pd2->o_desk);
edje_object_signal_emit(pd2->o_desk, "e,state,selected", "e");
}
else
{
if (pd2->current)
{
pd2->current = 0;
edje_object_signal_emit(pd2->o_desk, "e,state,unselected", "e");
}
}
}
}
static Pager_Desk *
_pager_desk_find(Pager *p, E_Desk *desk)
{
Eina_List *l;
Pager_Desk *pd;
EINA_LIST_FOREACH(p->desks, l, pd)
if (pd->desk == desk) return pd;
return NULL;
}
static void
_pager_desk_switch(Pager_Desk *pd1, Pager_Desk *pd2)
{
int c;
E_Zone *zone1, *zone2;
E_Desk *desk1, *desk2;
Pager_Win *pw;
Eina_List *l;
if ((!pd1) || (!pd2) || (!pd1->desk) || (!pd2->desk)) return;
if (pd1 == pd2) return;
desk1 = pd1->desk;
desk2 = pd2->desk;
zone1 = pd1->desk->zone;
zone2 = pd2->desk->zone;
/* Move opened windows from on desk to the other */
EINA_LIST_FOREACH(pd1->wins, l, pw)
{
if ((!pw) || (!pw->border) || (pw->border->iconic)) continue;
e_border_desk_set(pw->border, desk2);
}
EINA_LIST_FOREACH(pd2->wins, l, pw)
{
if ((!pw) || (!pw->border) || (pw->border->iconic)) continue;
e_border_desk_set(pw->border, desk1);
}
/* Modify desktop names in the config */
for (l = e_config->desktop_names, c = 0; l && c < 2; l = l->next)
{
E_Config_Desktop_Name *tmp_dn;
tmp_dn = l->data;
if (!tmp_dn) continue;
if ((tmp_dn->desk_x == desk1->x) &&
(tmp_dn->desk_y == desk1->y) &&
(tmp_dn->zone == (int) desk1->zone->num))
{
tmp_dn->desk_x = desk2->x;
tmp_dn->desk_y = desk2->y;
tmp_dn->zone = desk2->zone->num;
c++;
}
else if ((tmp_dn->desk_x == desk2->x) &&
(tmp_dn->desk_y == desk2->y) &&
(tmp_dn->zone == (int) desk2->zone->num))
{
tmp_dn->desk_x = desk1->x;
tmp_dn->desk_y = desk1->y;
tmp_dn->zone = desk1->zone->num;
c++;
}
}
if (c > 0) e_config_save();
e_desk_name_update();
/* Modify desktop backgrounds in the config */
for (l = e_config->desktop_backgrounds, c = 0; l && c < 2; l = l->next)
{
E_Config_Desktop_Background *tmp_db;
tmp_db = l->data;
if (!tmp_db) continue;
if ((tmp_db->desk_x == desk1->x) &&
(tmp_db->desk_y == desk1->y) &&
(tmp_db->zone == (int) desk1->zone->num))
{
tmp_db->desk_x = desk2->x;
tmp_db->desk_y = desk2->y;
tmp_db->zone = desk2->zone->num;
c++;
}
else if ((tmp_db->desk_x == desk2->x) &&
(tmp_db->desk_y == desk2->y) &&
(tmp_db->zone == (int) desk2->zone->num))
{
tmp_db->desk_x = desk1->x;
tmp_db->desk_y = desk1->y;
tmp_db->zone = desk1->zone->num;
c++;
}
}
if (c > 0) e_config_save();
/* If the current desktop has been switched, force to update of the screen */
if (desk2 == e_desk_current_get(zone2))
{
desk2->visible = 0;
e_desk_show(desk2);
}
if (desk1 == e_desk_current_get(zone1))
{
desk1->visible = 0;
e_desk_show(desk1);
}
}
static Pager_Win *
_pager_window_new(Pager_Desk *pd, E_Border *border)
{
Pager_Win *pw;
Evas_Object *o;
int visible;
if (!border) return NULL;
pw = E_NEW(Pager_Win, 1);
if (!pw) return NULL;
pw->border = border;
e_object_ref(E_OBJECT(border));
2005-10-11 10:22:39 -07:00
visible = ((!border->iconic) && (!border->client.netwm.state.skip_pager));
pw->skip_winlist = border->client.netwm.state.skip_pager;
pw->desk = pd;
o = edje_object_add(evas_object_evas_get(pd->pager->o_table));
pw->o_window = o;
e_theme_edje_object_set(o, "base/theme/modules/pager",
2006-08-22 21:02:03 -07:00
"e/modules/pager/window");
if (visible) evas_object_show(o);
e_layout_pack(pd->o_layout, pw->o_window);
e_layout_child_raise(pw->o_window);
evas_object_event_callback_add(o, EVAS_CALLBACK_MOUSE_DOWN,
_pager_window_cb_mouse_down, pw);
evas_object_event_callback_add(o, EVAS_CALLBACK_MOUSE_UP,
_pager_window_cb_mouse_up, pw);
evas_object_event_callback_add(o, EVAS_CALLBACK_MOUSE_MOVE,
_pager_window_cb_mouse_move, pw);
2005-10-11 10:22:39 -07:00
o = e_border_icon_add(border, evas_object_evas_get(pd->pager->o_table));
if (o)
{
pw->o_icon = o;
evas_object_show(o);
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
edje_object_part_swallow(pw->o_window, "e.swallow.icon", o);
}
if (border->client.icccm.urgent && !border->focused)
{
if (!(border->iconic))
edje_object_signal_emit(pd->o_desk, "e,state,urgent", "e");
edje_object_signal_emit(pw->o_window, "e,state,urgent", "e");
}
evas_object_show(o);
_pager_window_move(pw);
return pw;
}
static void
_pager_window_free(Pager_Win *pw)
{
if ((pw->drag.from_pager) && (pw->desk->pager->dragging))
pw->desk->pager->dragging = 0;
if (pw->o_window) evas_object_del(pw->o_window);
if (pw->o_icon) evas_object_del(pw->o_icon);
e_object_unref(E_OBJECT(pw->border));
E_FREE(pw);
}
static void
_pager_window_move(Pager_Win *pw)
{
e_layout_child_move(pw->o_window,
pw->border->x - pw->border->zone->x,
pw->border->y - pw->border->zone->y);
e_layout_child_resize(pw->o_window, pw->border->w, pw->border->h);
}
static Pager_Win *
_pager_window_find(Pager *p, E_Border *border)
{
Eina_List *l;
Pager_Desk *pd;
2005-10-11 10:22:39 -07:00
EINA_LIST_FOREACH(p->desks, l, pd)
{
Pager_Win *pw;
2005-10-11 10:22:39 -07:00
pw = _pager_desk_window_find(pd, border);
if (pw) return pw;
}
return NULL;
}
static Pager_Win *
_pager_desk_window_find(Pager_Desk *pd, E_Border *border)
{
Eina_List *l;
Pager_Win *pw;
2005-10-11 10:22:39 -07:00
EINA_LIST_FOREACH(pd->wins, l, pw)
if (pw->border == border) return pw;
2005-10-11 10:22:39 -07:00
return NULL;
}
static Pager_Popup *
_pager_popup_new(E_Zone *zone, int keyaction)
{
Pager_Popup *pp;
Evas_Coord w, h, zx, zy, zw, zh;
int x, y, height, width;
E_Desk *desk;
pp = E_NEW(Pager_Popup, 1);
if (!pp) return NULL;
/* Show popup */
pp->popup = e_popup_new(zone, 0, 0, 1, 1);
if (!pp->popup)
{
E_FREE(pp);
return NULL;
}
e_popup_layer_set(pp->popup, 255);
pp->pager = _pager_new(pp->popup->evas, zone);
pp->pager->popup = pp;
pp->urgent = 0;
e_zone_desk_count_get(zone, &x, &y);
if (keyaction)
height = pager_config->popup_act_height * y;
else
height = pager_config->popup_height * y;
width = height * (zone->w * x) / (zone->h * y);
evas_object_move(pp->pager->o_table, 0, 0);
evas_object_resize(pp->pager->o_table, width, height);
pp->o_bg = edje_object_add(pp->popup->evas);
e_theme_edje_object_set(pp->o_bg, "base/theme/modules/pager",
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
"e/modules/pager/popup");
desk = e_desk_current_get(zone);
if (desk)
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
edje_object_part_text_set(pp->o_bg, "e.text.label", desk->name);
evas_object_show(pp->o_bg);
edje_extern_object_min_size_set(pp->pager->o_table, width, height);
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
edje_object_part_swallow(pp->o_bg, "e.swallow.content", pp->pager->o_table);
edje_object_size_min_calc(pp->o_bg, &w, &h);
evas_object_move(pp->o_bg, 0, 0);
evas_object_resize(pp->o_bg, w, h);
e_popup_edje_bg_object_set(pp->popup, pp->o_bg);
//e_popup_ignore_events_set(pp->popup, 1);
e_zone_useful_geometry_get(zone, &zx, &zy, &zw, &zh);
zx -= zone->x;
zy -= zone->y;
e_popup_move_resize(pp->popup,
zx + ((zw - w) / 2), zy + ((zh - h) / 2), w, h);
e_bindings_mouse_grab(E_BINDING_CONTEXT_POPUP, pp->popup->evas_win);
e_bindings_wheel_grab(E_BINDING_CONTEXT_POPUP, pp->popup->evas_win);
e_popup_show(pp->popup);
pp->timer = NULL;
2006-05-31 09:19:11 -07:00
return pp;
}
static void
_pager_popup_free(Pager_Popup *pp)
{
if (pp->timer) ecore_timer_del(pp->timer);
evas_object_del(pp->o_bg);
_pager_free(pp->pager);
e_bindings_mouse_ungrab(E_BINDING_CONTEXT_POPUP, pp->popup->evas_win);
e_bindings_wheel_ungrab(E_BINDING_CONTEXT_POPUP, pp->popup->evas_win);
e_object_del(E_OBJECT(pp->popup));
E_FREE(pp);
}
static Pager_Popup *
_pager_popup_find(E_Zone *zone)
{
Eina_List *l;
Pager *p;
EINA_LIST_FOREACH(pagers, l, p)
if ((p->popup) && (p->zone == zone))
return p->popup;
return NULL;
}
static void
_pager_cb_obj_moveresize(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info __UNUSED__)
{
Instance *inst;
inst = data;
_pager_instance_drop_zone_recalc(inst);
}
2005-10-11 10:22:39 -07:00
static void
_button_cb_mouse_down(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info)
{
Instance *inst;
Evas_Event_Mouse_Down *ev;
inst = data;
ev = event_info;
if ((ev->button == 3) && (!pager_config->menu))
{
E_Menu *m;
E_Menu_Item *mi;
int cx, cy;
m = e_menu_new();
mi = e_menu_item_new(m);
e_menu_item_label_set(mi, _("Settings"));
e_util_menu_item_theme_icon_set(mi, "configure");
e_menu_item_callback_set(mi, _pager_inst_cb_menu_configure, NULL);
m = e_gadcon_client_util_menu_items_append(inst->gcc, m, 0);
e_menu_post_deactivate_callback_set(m, _menu_cb_post, inst);
pager_config->menu = m;
if (e_configure_registry_exists("screen/virtual_desktops"))
{
mi = e_menu_item_new_relative(m, NULL);
e_menu_item_label_set(mi, _("Virtual Desktops Settings"));
e_util_menu_item_theme_icon_set(mi, "preferences-desktop");
e_menu_item_callback_set(mi, _pager_inst_cb_menu_virtual_desktops_dialog, inst);
}
e_gadcon_canvas_zone_geometry_get(inst->gcc->gadcon, &cx, &cy,
NULL, NULL);
e_menu_activate_mouse(m,
e_util_zone_current_get(e_manager_current_get()),
cx + ev->output.x, cy + ev->output.y, 1, 1,
E_MENU_POP_DIRECTION_DOWN, ev->timestamp);
}
}
static void
_menu_cb_post(void *data __UNUSED__, E_Menu *m __UNUSED__)
{
if (!pager_config->menu) return;
e_object_del(E_OBJECT(pager_config->menu));
pager_config->menu = NULL;
}
static void
_pager_inst_cb_menu_configure(void *data __UNUSED__, E_Menu *m __UNUSED__, E_Menu_Item *mi __UNUSED__)
{
if (!pager_config) return;
if (pager_config->config_dialog) return;
/* FIXME: pass zone config item */
_config_pager_module(NULL);
}
static E_Config_Dialog *
_pager_config_dialog(E_Container *con __UNUSED__, const char *params __UNUSED__)
{
if (!pager_config) return NULL;
if (pager_config->config_dialog) return NULL;
/* FIXME: pass zone config item */
_config_pager_module(NULL);
return pager_config->config_dialog;
}
static void
_pager_inst_cb_menu_virtual_desktops_dialog(void *data, E_Menu *m __UNUSED__, E_Menu_Item *mi __UNUSED__)
{
Instance *inst;
inst = data;
e_configure_registry_call("screen/virtual_desktops",
inst->gcc->gadcon->zone->container, NULL);
}
static void
_pager_instance_drop_zone_recalc(Instance *inst)
{
Evas_Coord x, y, w, h;
e_gadcon_client_viewport_geometry_get(inst->gcc, &x, &y, &w, &h);
e_drop_handler_geometry_set(inst->pager->drop_handler, x, y, w, h);
}
void
_pager_cb_config_updated(void)
{
if (!pager_config) return;
}
static Eina_Bool
_pager_cb_event_border_resize(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Border_Resize *ev = event;
Eina_List *l, *l2;
Pager *p;
EINA_LIST_FOREACH(pagers, l, p)
{
Pager_Desk *pd;
if (p->zone != ev->border->zone) continue;
EINA_LIST_FOREACH(p->desks, l2, pd)
{
Pager_Win *pw;
pw = _pager_desk_window_find(pd, ev->border);
if (pw) _pager_window_move(pw);
}
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_border_move(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Border_Move *ev = event;
Eina_List *l;
Pager_Win *pw;
Pager_Desk *pd;
Pager *p;
EINA_LIST_FOREACH(pagers, l, p)
{
Eina_List *l2;
if (p->zone != ev->border->zone) continue;
EINA_LIST_FOREACH(p->desks, l2, pd)
{
pw = _pager_desk_window_find(pd, ev->border);
if (pw) _pager_window_move(pw);
}
}
if ((act_popup) && (act_popup->pager->zone == ev->border->zone))
{
EINA_LIST_FOREACH(act_popup->pager->desks, l, pd)
{
pw = _pager_desk_window_find(pd, ev->border);
if (pw) _pager_window_move(pw);
}
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_border_add(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Border_Add *ev = event;
Eina_List *l;
Pager *p;
EINA_LIST_FOREACH(pagers, l, p)
{
Pager_Desk *pd;
Pager_Win *pw;
if ((p->zone != ev->border->zone) ||
(_pager_window_find(p, ev->border)))
continue;
pd = _pager_desk_find(p, ev->border->desk);
if (!pd) continue;
pw = _pager_window_new(pd, ev->border);
if (pw) pd->wins = eina_list_append(pd->wins, pw);
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_border_remove(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Border_Remove *ev = event;
Eina_List *l;
Pager *p;
EINA_LIST_FOREACH(pagers, l, p)
{
Eina_List *l2;
Pager_Desk *pd;
if (p->zone != ev->border->zone) continue;
EINA_LIST_FOREACH(p->desks, l2, pd)
2005-02-10 07:21:23 -08:00
{
Pager_Win *pw;
pw = _pager_desk_window_find(pd, ev->border);
if (!pw) continue;
pd->wins = eina_list_remove(pd->wins, pw);
_pager_window_free(pw);
2005-02-10 07:21:23 -08:00
}
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_border_iconify(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Border_Iconify *ev = event;
Eina_List *l;
Pager *p;
EINA_LIST_FOREACH(pagers, l, p)
{
Eina_List *l2;
Pager_Desk *pd;
if (p->zone != ev->border->zone) continue;
EINA_LIST_FOREACH(p->desks, l2, pd)
{
Pager_Win *pw;
pw = _pager_desk_window_find(pd, ev->border);
if (!pw) continue;
if ((pw->drag.from_pager) && (pw->desk->pager->dragging))
pw->desk->pager->dragging = 0;
evas_object_hide(pw->o_window);
}
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_border_uniconify(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Border_Uniconify *ev = event;
Eina_List *l;
Pager *p;
EINA_LIST_FOREACH(pagers, l, p)
{
Eina_List *l2;
Pager_Desk *pd;
if (p->zone != ev->border->zone) continue;
EINA_LIST_FOREACH(p->desks, l2, pd)
{
Pager_Win *pw;
pw = _pager_desk_window_find(pd, ev->border);
if ((pw) && (!pw->skip_winlist)) evas_object_show(pw->o_window);
}
2005-02-10 07:21:23 -08:00
}
return ECORE_CALLBACK_PASS_ON;
2005-02-10 07:21:23 -08:00
}
static Eina_Bool
_pager_cb_event_border_stick(void *data __UNUSED__, int type __UNUSED__, void *event)
2005-02-10 07:21:23 -08:00
{
E_Event_Border_Stick *ev = event;
Eina_List *l;
Pager *p;
2005-02-10 07:21:23 -08:00
EINA_LIST_FOREACH(pagers, l, p)
2005-02-10 07:21:23 -08:00
{
Eina_List *l2;
Pager_Desk *pd;
Pager_Win *pw;
if (p->zone != ev->border->zone) continue;
pw = _pager_window_find(p, ev->border);
if (!pw) continue;
EINA_LIST_FOREACH(p->desks, l2, pd)
if (ev->border->desk != pd->desk)
{
pw = _pager_window_new(pd, ev->border);
if (pw) pd->wins = eina_list_append(pd->wins, pw);
}
2005-02-10 07:21:23 -08:00
}
return ECORE_CALLBACK_PASS_ON;
2005-02-10 07:21:23 -08:00
}
static Eina_Bool
_pager_cb_event_border_unstick(void *data __UNUSED__, int type __UNUSED__, void *event)
2005-02-10 07:21:23 -08:00
{
E_Event_Border_Unstick *ev = event;
Eina_List *l;
Pager *p;
2005-02-10 07:21:23 -08:00
EINA_LIST_FOREACH(pagers, l, p)
2005-02-10 07:21:23 -08:00
{
Pager_Desk *pd;
Eina_List *l2;
if (p->zone != ev->border->zone) continue;
EINA_LIST_FOREACH(p->desks, l2, pd)
if (ev->border->desk != pd->desk)
{
Pager_Win *pw;
pw = _pager_desk_window_find(pd, ev->border);
if (pw)
{
pd->wins = eina_list_remove(pd->wins, pw);
_pager_window_free(pw);
}
}
}
return ECORE_CALLBACK_PASS_ON;
}
static void
_pager_window_desk_change(Pager *pager, E_Border *bd)
{
Eina_List *l;
Pager_Win *pw;
Pager_Desk *pd;
2005-10-11 10:22:39 -07:00
/* if this pager is not for the zone of the border */
if (pager->zone != bd->zone)
{
/* look at all desks in the pager */
EINA_LIST_FOREACH(pager->desks, l, pd)
{
/* find this border in this desk */
pw = _pager_desk_window_find(pd, bd);
if (!pw) continue;
/* if it is found - remove it. it does not belong in this
* pager as it probably moves zones */
pd->wins = eina_list_remove(pd->wins, pw);
_pager_window_free(pw);
}
return;
}
/* and this pager zone is for this border */
/* see if the window is in this pager at all */
pw = _pager_window_find(pager, bd);
if (pw)
{
/* is it sticky */
if (bd->sticky)
{
/* if its sticky and in this pager - its already everywhere, so abort
* doing anything else */
return;
}
/* move it to the right desk */
/* find the pager desk of the target desk */
pd = _pager_desk_find(pager, bd->desk);
if (pd)
{
Pager_Win *pw2 = NULL;
E_Border *bd_above;
/* remove it from whatever desk it was on */
pw->desk->wins = eina_list_remove(pw->desk->wins, pw);
e_layout_unpack(pw->o_window);
/* add it to the one its MEANT to be on */
pw->desk = pd;
pd->wins = eina_list_append(pd->wins, pw);
e_layout_pack(pd->o_layout, pw->o_window);
bd_above = e_util_desk_border_above(pw->border);
if (bd_above)
pw2 = _pager_desk_window_find(pd, bd_above);
if (pw2)
e_layout_child_lower_below(pw->o_window, pw2->o_window);
else
e_layout_child_raise(pw->o_window);
_pager_window_move(pw);
}
}
Fix common misspellings Some misspellings were manually reverted since in E there are tons of .po with phrases in other languages. Other than that all the changes in the following files were reverted: * src/modules/illume/dicts/English_(US).dic * src/modules/illume-keyboard/dicts/English_(US).dic Following misspellings were fixed: acquited->acquitted adres->address adress->address alreayd->already aquire->acquire arbitarily->arbitrarily cant->can't Capetown->Cape carefull->careful causalities->casualties Celcius->Celsius certian->certain commandoes->commandos considerd->considered conveyer->conveyor dependant->dependent didnt->didn't discontentment->discontent doesnt->doesn't everytime->every exemple->example existance->existence existant->existent existince->existence Farenheit->Fahrenheit forbad->forbade funguses->fungi guage->gauge guerilla->guerrilla guerillas->guerrillas happend->happened hasnt->hasn't heros->heroes inbetween->between independant->independent inital->initial intrusted->entrusted irregardless->regardless isnt->isn't knifes->knives layed->laid loosing->losing marrage->marriage midwifes->midwives miniscule->minuscule monickers->monikers mroe->more noone->no one occured->occurred omre->more paralell->parallel payed->paid planed->planned quitted->quit quizes->quizzes seperated->separated seperate->separate shoudl->should similiar->similar simplier->simpler specifiying->specifying teh->the toke->took torpedos->torpedoes Tuscon->Tucson unecessary->unnecessary useage->usage usefull->useful useing->using waht->what wanna->want whith->with wich->which withing->within SVN revision: 52006
2010-09-08 16:59:07 -07:00
/* the border isn't in this pager at all - it must have moved zones */
else
{
if (!bd->sticky)
{
/* find the pager desk it needs to go to */
pd = _pager_desk_find(pager, bd->desk);
if (pd)
{
/* create it and add it */
pw = _pager_window_new(pd, bd);
if (pw)
{
Pager_Win *pw2 = NULL;
E_Border *bd_above;
pd->wins = eina_list_append(pd->wins, pw);
bd_above = e_util_desk_border_above(pw->border);
if (bd_above)
pw2 = _pager_desk_window_find(pd, bd_above);
if (pw2)
e_layout_child_lower_below(pw->o_window, pw2->o_window);
else
e_layout_child_raise(pw->o_window);
_pager_window_move(pw);
}
}
}
else
{
/* go through all desks */
EINA_LIST_FOREACH(pager->desks, l, pd)
{
/* create it and add it */
pw = _pager_window_new(pd, bd);
if (pw)
{
Pager_Win *pw2 = NULL;
E_Border *bd_above;
pd->wins = eina_list_append(pd->wins, pw);
bd_above = e_util_desk_border_above(pw->border);
if (bd_above)
pw2 = _pager_desk_window_find(pd, bd_above);
if (pw2)
e_layout_child_lower_below(pw->o_window, pw2->o_window);
else
e_layout_child_raise(pw->o_window);
_pager_window_move(pw);
}
}
}
}
}
static Eina_Bool
_pager_cb_event_border_desk_set(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Border_Desk_Set *ev = event;
Eina_List *l;
Pager *p;
EINA_LIST_FOREACH(pagers, l, p)
_pager_window_desk_change(p, ev->border);
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_border_stack(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Border_Stack *ev = event;
Eina_List *l;
Pager *p;
EINA_LIST_FOREACH(pagers, l, p)
{
Eina_List *l2;
Pager_Desk *pd;
if (p->zone != ev->border->zone) continue;
EINA_LIST_FOREACH(p->desks, l2, pd)
{
Pager_Win *pw, *pw2 = NULL;
pw = _pager_desk_window_find(pd, ev->border);
if (pw)
{
if (ev->stack)
{
pw2 = _pager_desk_window_find(pd, ev->stack);
if (!pw2)
{
/* This border is on another desk... */
E_Border *bd = NULL;
if (ev->type == E_STACKING_ABOVE)
bd = e_util_desk_border_below(ev->border);
else if (ev->type == E_STACKING_BELOW)
bd = e_util_desk_border_above(ev->border);
if (bd) pw2 = _pager_desk_window_find(pd, bd);
}
}
if (ev->type == E_STACKING_ABOVE)
{
if (pw2)
e_layout_child_raise_above(pw->o_window, pw2->o_window);
else
/* If we aren't above any window, we are at the bottom */
e_layout_child_lower(pw->o_window);
}
else if (ev->type == E_STACKING_BELOW)
{
if (pw2)
e_layout_child_lower_below(pw->o_window, pw2->o_window);
else
/* If we aren't below any window, we are at the top */
e_layout_child_raise(pw->o_window);
}
}
}
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_border_icon_change(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Border_Icon_Change *ev = event;
Eina_List *l;
Pager *p;
EINA_LIST_FOREACH(pagers, l, p)
{
Eina_List *l2;
Pager_Desk *pd;
if (p->zone != ev->border->zone) continue;
EINA_LIST_FOREACH(p->desks, l2, pd)
{
Pager_Win *pw;
pw = _pager_desk_window_find(pd, ev->border);
if (pw)
{
Evas_Object *o;
if (pw->o_icon)
{
evas_object_del(pw->o_icon);
pw->o_icon = NULL;
}
o = e_border_icon_add(ev->border,
evas_object_evas_get(p->o_table));
if (o)
{
pw->o_icon = o;
evas_object_show(o);
edje_object_part_swallow(pw->o_window,
"e.swallow.icon", o);
}
}
}
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_border_urgent_change(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Border_Urgent_Change *ev;
Eina_List *l, *l2;
Pager_Popup *pp;
E_Zone *zone;
int urgent;
Pager *p;
Pager_Desk *pd;
Pager_Win *pw;
ev = event;
zone = ev->border->zone;
urgent = ev->border->client.icccm.urgent;
if (pager_config->popup_urgent && (pager_config->popup_urgent_focus ||
(!pager_config->popup_urgent_focus && !ev->border->focused)))
{
pp = _pager_popup_find(zone);
if ((!pp) && (urgent) && !(ev->border->iconic))
{
pp = _pager_popup_new(zone, 0);
if ((pp) && (!pager_config->popup_urgent_stick))
pp->timer = ecore_timer_add(pager_config->popup_urgent_speed,
_pager_popup_cb_timeout, pp);
if (pp) pp->urgent = 1;
}
}
EINA_LIST_FOREACH(pagers, l, p)
{
if (p->zone != zone) continue;
EINA_LIST_FOREACH(p->desks, l2, pd)
{
pw = _pager_desk_window_find(pd, ev->border);
if (pw)
{
if (urgent && !ev->border->focused)
{
if (!(ev->border->iconic))
{
if ((pd->pager) && (pd->pager->inst) &&
(!pager_config->popup_urgent))
e_gadcon_urgent_show(pd->pager->inst->gcc->gadcon);
edje_object_signal_emit(pd->o_desk,
"e,state,urgent", "e");
}
edje_object_signal_emit(pw->o_window,
"e,state,urgent", "e");
}
else
{
if (!(ev->border->iconic))
edje_object_signal_emit(pd->o_desk,
"e,state,not_urgent", "e");
edje_object_signal_emit(pw->o_window,
"e,state,not_urgent", "e");
}
}
}
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_border_focus_in(void *data __UNUSED__, int type __UNUSED__, void *event)
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
{
E_Event_Border_Focus_In *ev;
Instance *inst;
Eina_List *l, *l2;
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
Pager_Popup *pp;
Pager_Desk *pd;
Pager_Win *pw;
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
E_Zone *zone;
ev = event;
zone = ev->border->zone;
EINA_LIST_FOREACH(pager_config->instances, l, inst)
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
{
if (inst->pager->zone != zone) continue;
EINA_LIST_FOREACH(inst->pager->desks, l2, pd)
{
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
pw = _pager_desk_window_find(pd, ev->border);
if (pw)
{
edje_object_signal_emit(pw->o_window,
"e,state,focused", "e");
break;
}
}
}
pp = _pager_popup_find(zone);
if (!pp) return ECORE_CALLBACK_PASS_ON;
EINA_LIST_FOREACH(pp->pager->desks, l, pd)
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
{
pw = _pager_desk_window_find(pd, ev->border);
if (pw)
{
edje_object_signal_emit(pw->o_window,
"e,state,focused", "e");
break;
}
}
return ECORE_CALLBACK_PASS_ON;
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
}
static Eina_Bool
_pager_cb_event_border_focus_out(void *data __UNUSED__, int type __UNUSED__, void *event)
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
{
E_Event_Border_Focus_Out *ev;
Eina_List *l;
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
Pager_Popup *pp;
Pager_Desk *pd;
Pager_Win *pw;
Instance *inst;
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
E_Zone *zone;
ev = event;
zone = ev->border->zone;
EINA_LIST_FOREACH(pager_config->instances, l, inst)
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
{
Eina_List *l2;
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
if (inst->pager->zone != zone) continue;
EINA_LIST_FOREACH(inst->pager->desks, l2, pd)
{
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
pw = _pager_desk_window_find(pd, ev->border);
if (pw)
{
edje_object_signal_emit(pw->o_window,
"e,state,unfocused", "e");
break;
}
}
}
pp = _pager_popup_find(zone);
if (!pp) return ECORE_CALLBACK_PASS_ON;
EINA_LIST_FOREACH(pp->pager->desks, l, pd)
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
{
pw = _pager_desk_window_find(pd, ev->border);
if (pw)
{
edje_object_signal_emit(pw->o_window,
"e,state,unfocused", "e");
break;
}
}
return ECORE_CALLBACK_PASS_ON;
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
}
static Eina_Bool
_pager_cb_event_border_property(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Border_Property *ev = event;
Eina_List *l, *l2;
int found = 0;
Pager *p;
Pager_Win *pw;
Pager_Desk *pd;
EINA_LIST_FOREACH(pagers, l, p)
{
if (p->zone != ev->border->zone) continue;
EINA_LIST_FOREACH(p->desks, l2, pd)
{
pw = _pager_desk_window_find(pd, ev->border);
if (pw)
{
found = 1;
if (ev->border->client.netwm.state.skip_pager)
{
pd->wins = eina_list_remove(pd->wins, pw);
_pager_window_free(pw);
}
}
}
}
if (found) return ECORE_CALLBACK_PASS_ON;
/* If we did not find this window in the pager, then add it because
* the skip_pager state may have changed to 1 */
EINA_LIST_FOREACH(pagers, l, p)
{
if ((p->zone != ev->border->zone) ||
(_pager_window_find(p, ev->border)))
continue;
if (!ev->border->sticky)
{
pd = _pager_desk_find(p, ev->border->desk);
if (pd)
{
pw = _pager_window_new(pd, ev->border);
if (pw)
{
Pager_Win *pw2 = NULL;
E_Border *bd;
pd->wins = eina_list_append(pd->wins, pw);
bd = e_util_desk_border_above(pw->border);
if (bd)
pw2 = _pager_desk_window_find(pd, bd);
if (pw2)
e_layout_child_lower_below(pw->o_window, pw2->o_window);
else
e_layout_child_raise(pw->o_window);
_pager_window_move(pw);
}
}
}
else
{
EINA_LIST_FOREACH(p->desks, l2, pd)
{
pw = _pager_window_new(pd, ev->border);
if (pw)
{
Pager_Win *pw2 = NULL;
E_Border *bd;
pd->wins = eina_list_append(pd->wins, pw);
bd = e_util_desk_border_above(pw->border);
if (bd)
pw2 = _pager_desk_window_find(pd, bd);
if (pw2)
e_layout_child_lower_below(pw->o_window, pw2->o_window);
else
e_layout_child_raise(pw->o_window);
_pager_window_move(pw);
}
}
}
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_zone_desk_count_set(void *data __UNUSED__, int type __UNUSED__, void *event __UNUSED__)
{
Eina_List *l;
Pager *p;
EINA_LIST_FOREACH(pagers, l, p)
{
_pager_empty(p);
_pager_fill(p);
if (p->inst) _gc_orient(p->inst->gcc, p->inst->gcc->gadcon->orient);
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_desk_show(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Desk_Show *ev = event;
Eina_List *l;
Pager *p;
Pager_Popup *pp;
Pager_Desk *pd;
EINA_LIST_FOREACH(pagers, l, p)
{
if (p->zone != ev->desk->zone) continue;
pd = _pager_desk_find(p, ev->desk);
if (pd) _pager_desk_select(pd);
if (p->popup)
edje_object_part_text_set(p->popup->o_bg, "text", ev->desk->name);
}
if ((pager_config->popup) && (!act_popup))
{
if ((pp = _pager_popup_find(ev->desk->zone)))
ecore_timer_del(pp->timer);
else
pp = _pager_popup_new(ev->desk->zone, 0);
if (pp)
{
pp->timer = ecore_timer_add(pager_config->popup_speed,
_pager_popup_cb_timeout, pp);
pd = _pager_desk_find(pp->pager, ev->desk);
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
if (pd)
{
_pager_desk_select(pd);
edje_object_part_text_set(pp->o_bg, "e.text.label",
ev->desk->name);
}
ecore_timer_del(pp->timer);
pp->timer = ecore_timer_add(pager_config->popup_speed,
_pager_popup_cb_timeout, pp);
}
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_desk_name_change(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Desk_Name_Change *ev = event;
Eina_List *l;
Pager *p;
2005-10-11 10:22:39 -07:00
EINA_LIST_FOREACH(pagers, l, p)
{
Pager_Desk *pd;
if (p->zone != ev->desk->zone) continue;
pd = _pager_desk_find(p, ev->desk);
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
if (pager_config->show_desk_names)
{
if (pd)
edje_object_part_text_set(pd->o_desk, "e.text.label",
ev->desk->name);
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
}
else
{
if (pd)
edje_object_part_text_set(pd->o_desk, "e.text.label", "");
}
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_cb_event_container_resize(void *data __UNUSED__, int type __UNUSED__, void *event)
{
E_Event_Container_Resize *ev = event;
Eina_List *l;
Pager *p;
EINA_LIST_FOREACH(pagers, l, p)
{
Eina_List *l2;
Pager_Desk *pd;
if (p->zone->container != ev->container) continue;
EINA_LIST_FOREACH(p->desks, l2, pd)
e_layout_virtual_size_set(pd->o_layout, pd->desk->zone->w,
pd->desk->zone->h);
if (p->inst) _gc_orient(p->inst->gcc, p->inst->gcc->gadcon->orient);
/* TODO if (p->popup) */
}
return ECORE_CALLBACK_PASS_ON;
}
static void
_pager_window_cb_mouse_down(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info)
{
Evas_Event_Mouse_Down *ev;
Pager_Win *pw;
2005-10-11 10:22:39 -07:00
ev = event_info;
pw = data;
if (!pw) return;
if (pw->desk->pager->popup && !act_popup) return;
if (!pw->desk->pager->popup && ev->button == 3) return;
if (ev->button == (int) pager_config->btn_desk) return;
if ((ev->button == (int) pager_config->btn_drag) ||
(ev->button == (int) pager_config->btn_noplace))
{
Evas_Coord ox, oy;
evas_object_geometry_get(pw->o_window, &ox, &oy, NULL, NULL);
pw->drag.in_pager = 1;
pw->drag.x = ev->canvas.x;
pw->drag.y = ev->canvas.y;
pw->drag.dx = ox - ev->canvas.x;
pw->drag.dy = oy - ev->canvas.y;
pw->drag.start = 1;
pw->drag.button = ev->button;
#if 0
/* FIXME: disable move in pager for now, as dropping in between
desks causes lost window */
pw->drag.no_place = 0;
if (ev->button == pager_config->btn_noplace) pw->drag.no_place = 1;
#else
pw->drag.no_place = 1;
#endif
}
}
static void
_pager_window_cb_mouse_up(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info)
{
Evas_Event_Mouse_Up *ev;
Pager_Win *pw;
Pager *p;
ev = event_info;
pw = data;
if (!pw) return;
p = pw->desk->pager;
if (pw->desk->pager->popup && !act_popup) return;
if (ev->button == (int) pager_config->btn_desk) return;
if ((ev->button == (int) pager_config->btn_drag) ||
(ev->button == (int) pager_config->btn_noplace))
{
if (!pw->drag.from_pager)
{
if (!pw->drag.start) p->just_dragged = 1;
pw->drag.in_pager = 0;
pw->drag.start = 0;
p->dragging = 0;
}
}
}
static void
_pager_window_cb_mouse_move(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info)
{
Evas_Event_Mouse_Move *ev;
Pager_Win *pw;
E_Drag *drag;
Evas_Object *o, *oo;
Evas_Coord x, y, w, h;
const char *drag_types[] =
{ "enlightenment/pager_win", "enlightenment/border" };
Evas_Coord dx, dy;
unsigned int resist = 0;
Evas_Coord mx, my, vx, vy;
Pager_Desk *pd;
ev = event_info;
pw = data;
if (!pw) return;
if (pw->border->lock_user_location) return;
if ((pw->desk->pager->popup) && (!act_popup)) return;
/* prevent drag for a few pixels */
if (pw->drag.start)
{
dx = pw->drag.x - ev->cur.output.x;
dy = pw->drag.y - ev->cur.output.y;
if ((pw->desk) && (pw->desk->pager))
resist = pager_config->drag_resist;
if (((unsigned int) (dx * dx) + (unsigned int) (dy * dy)) <=
(resist * resist)) return;
pw->desk->pager->dragging = 1;
pw->drag.start = 0;
}
/* dragging this win around inside the pager */
if (pw->drag.in_pager)
{
/* m for mouse */
mx = ev->cur.canvas.x;
my = ev->cur.canvas.y;
/* find desk at pointer */
pd = _pager_desk_at_coord(pw->desk->pager, mx, my);
if ((pd) && (!pw->drag.no_place))
{
int zx, zy;
e_zone_useful_geometry_get(pd->desk->zone, &zx, &zy, NULL, NULL);
e_layout_coord_canvas_to_virtual(pd->o_layout,
mx + pw->drag.dx,
my + pw->drag.dy, &vx, &vy);
if (pd != pw->desk)
e_border_desk_set(pw->border, pd->desk);
e_border_move(pw->border, vx + zx, vy + zy);
}
else
{
evas_object_geometry_get(pw->o_window, &x, &y, &w, &h);
evas_object_hide(pw->o_window);
drag = e_drag_new(pw->desk->pager->zone->container,
x, y, drag_types, 2, pw, -1,
_pager_window_cb_drag_convert,
_pager_window_cb_drag_finished);
o = edje_object_add(drag->evas);
e_theme_edje_object_set(o, "base/theme/modules/pager",
"e/modules/pager/window");
evas_object_show(o);
oo = e_border_icon_add(pw->border, drag->evas);
if (oo)
{
evas_object_show(oo);
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
edje_object_part_swallow(o, "e.swallow.icon", oo);
}
e_drag_object_set(drag, o);
e_drag_resize(drag, w, h);
e_drag_start(drag, x - pw->drag.dx, y - pw->drag.dy);
/* this prevents the desk from switching on drags */
pw->drag.from_pager = pw->desk->pager;
pw->drag.from_pager->dragging = 1;
pw->drag.in_pager = 0;
}
}
}
static void *
_pager_window_cb_drag_convert(E_Drag *drag, const char *type)
{
Pager_Win *pw;
pw = drag->data;
if (!strcmp(type, "enlightenment/pager_win")) return pw;
if (!strcmp(type, "enlightenment/border")) return pw->border;
return NULL;
}
static void
_pager_window_cb_drag_finished(E_Drag *drag, int dropped)
{
Pager_Win *pw;
E_Container *cont;
E_Zone *zone;
E_Desk *desk;
int x, y, dx, dy;
pw = drag->data;
if (!pw) return;
evas_object_show(pw->o_window);
if (!dropped)
{
int zx, zy, zw, zh;
/* wasn't dropped (on pager). move it to position of mouse on screen */
cont = e_container_current_get(e_manager_current_get());
zone = e_zone_current_get(cont);
desk = e_desk_current_get(zone);
e_border_zone_set(pw->border, zone);
e_border_desk_set(pw->border, desk);
ecore_x_pointer_last_xy_get(&x, &y);
dx = (pw->border->w / 2);
dy = (pw->border->h / 2);
e_zone_useful_geometry_get(zone, &zx, &zy, &zw, &zh);
/* offset so that center of window is on mouse, but keep within desk bounds */
if (dx < x)
{
x -= dx;
if ((pw->border->w < zw) &&
(x + pw->border->w > zx + zw))
x -= x + pw->border->w - (zx + zw);
}
else x = 0;
if (dy < y)
{
y -= dy;
if ((pw->border->h < zh) &&
(y + pw->border->h > zy + zh))
y -= y + pw->border->h - (zy + zh);
}
else y = 0;
e_border_move(pw->border, x, y);
if (!(pw->border->lock_user_stacking)) e_border_raise(pw->border);
}
if (pw->drag.from_pager) pw->drag.from_pager->dragging = 0;
pw->drag.from_pager = NULL;
if (act_popup)
{
e_grabinput_get(input_window, 0, input_window);
if (!hold_count) _pager_popup_hide(1);
}
}
static void
_pager_inst_cb_scroll(void *data)
{
Pager *p;
2005-04-25 11:02:26 -07:00
p = data;
_pager_update_drop_position(p, p->dnd_x, p->dnd_y);
}
static void
_pager_update_drop_position(Pager *p, Evas_Coord x, Evas_Coord y)
{
Pager_Desk *pd, *pd2;
Eina_List *l;
p->dnd_x = x;
p->dnd_y = y;
pd = _pager_desk_at_coord(p, x, y);
if (pd == p->active_drop_pd) return;
for (l = p->desks; l; l = l->next)
{
pd2 = l->data;
if (pd == pd2)
edje_object_signal_emit(pd2->o_desk, "e,action,drag,in", "e");
else if (pd2 == p->active_drop_pd)
edje_object_signal_emit(pd2->o_desk, "e,action,drag,out", "e");
}
p->active_drop_pd = pd;
}
static void
_pager_drop_cb_enter(void *data, const char *type __UNUSED__, void *event_info)
{
E_Event_Dnd_Enter *ev;
Pager *p;
ev = event_info;
p = data;
/* FIXME this fixes a segv, but the case is not easy
* reproduceable. this makes no sense either since
* the same 'pager' is passed to e_drop_handler_add
* and it works without this almost all the time.
* so this must be an issue with e_dnd code... i guess */
if (act_popup) p = act_popup->pager;
_pager_update_drop_position(p, ev->x, ev->y);
if (p->inst)
{
e_gadcon_client_autoscroll_cb_set(p->inst->gcc, _pager_inst_cb_scroll, p);
e_gadcon_client_autoscroll_update(p->inst->gcc, ev->x, ev->y);
}
}
static void
_pager_drop_cb_move(void *data, const char *type __UNUSED__, void *event_info)
{
E_Event_Dnd_Move *ev;
Pager *p;
ev = event_info;
p = data;
if (act_popup) p = act_popup->pager;
_pager_update_drop_position(p, ev->x, ev->y);
if (p->inst)
e_gadcon_client_autoscroll_update(p->inst->gcc, ev->x, ev->y);
}
static void
_pager_drop_cb_leave(void *data, const char *type __UNUSED__, void *event_info __UNUSED__)
{
Pager *p;
Eina_List *l;
p = data;
if (act_popup) p = act_popup->pager;
for (l = p->desks; l && p->active_drop_pd; l = l->next)
{
Pager_Desk *pd;
pd = l->data;
if (pd == p->active_drop_pd)
{
edje_object_signal_emit(pd->o_desk, "e,action,drag,out", "e");
p->active_drop_pd = NULL;
}
}
if (p->inst) e_gadcon_client_autoscroll_cb_set(p->inst->gcc, NULL, NULL);
}
static void
_pager_drop_cb_drop(void *data, const char *type, void *event_info)
{
E_Event_Dnd_Drop *ev;
Pager_Desk *pd;
Pager_Desk *pd2 = NULL;
E_Border *bd = NULL;
Eina_List *l;
int dx = 0, dy = 0;
Pager_Win *pw = NULL;
Evas_Coord wx, wy, wx2, wy2;
Evas_Coord nx, ny;
Pager *p;
ev = event_info;
p = data;
if (act_popup) p = act_popup->pager;
pd = _pager_desk_at_coord(p, ev->x, ev->y);
if (pd)
{
if (!strcmp(type, "enlightenment/pager_win"))
{
pw = (Pager_Win *)(ev->data);
if (pw)
{
bd = pw->border;
dx = pw->drag.dx;
dy = pw->drag.dy;
}
}
else if (!strcmp(type, "enlightenment/border"))
{
bd = ev->data;
e_layout_coord_virtual_to_canvas(pd->o_layout, bd->x, bd->y,
&wx, &wy);
e_layout_coord_virtual_to_canvas(pd->o_layout, bd->x + bd->w,
bd->y + bd->h, &wx2, &wy2);
dx = (wx - wx2) / 2;
dy = (wy - wy2) / 2;
}
else if (!strcmp(type, "enlightenment/vdesktop"))
{
pd2 = ev->data;
if (!pd2) return;
_pager_desk_switch(pd, pd2);
}
else
return;
if (bd)
{
if (bd->iconic) e_border_uniconify(bd);
e_border_desk_set(bd, pd->desk);
if ((!pw) || ((pw) && (!pw->drag.no_place)))
{
int zx, zy;
e_layout_coord_canvas_to_virtual(pd->o_layout,
ev->x + dx,
ev->y + dy,
&nx, &ny);
e_zone_useful_geometry_get(pd->desk->zone,
&zx, &zy, NULL, NULL);
e_border_move(bd, nx + zx, ny + zy);
}
}
}
for (l = p->desks; l && p->active_drop_pd; l = l->next)
{
pd = l->data;
if (pd == p->active_drop_pd)
{
edje_object_signal_emit(pd->o_desk, "e,action,drag,out", "e");
p->active_drop_pd = NULL;
}
}
if (p->inst) e_gadcon_client_autoscroll_cb_set(p->inst->gcc, NULL, NULL);
}
static void
_pager_desk_cb_mouse_down(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info)
{
Evas_Event_Mouse_Down *ev;
Pager_Desk *pd;
Evas_Coord ox, oy;
ev = event_info;
pd = data;
if (!pd) return;
if ((!pd->pager->popup) && (ev->button == 3)) return;
if (ev->button == (int) pager_config->btn_desk)
{
evas_object_geometry_get(pd->o_desk, &ox, &oy, NULL, NULL);
pd->drag.start = 1;
pd->drag.in_pager = 1;
pd->drag.dx = ox - ev->canvas.x;
pd->drag.dy = oy - ev->canvas.y;
pd->drag.x = ev->canvas.x;
pd->drag.y = ev->canvas.y;
pd->drag.button = ev->button;
}
}
2005-09-20 10:09:27 -07:00
static void
_pager_desk_cb_mouse_up(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info)
2005-09-20 10:09:27 -07:00
{
Evas_Event_Mouse_Up *ev;
Pager_Desk *pd;
Pager *p;
2005-09-20 10:09:27 -07:00
ev = event_info;
pd = data;
if (!pd) return;
p = pd->pager;
/* FIXME: pd->pager->dragging is 0 when finishing a drag from desk to desk */
if ((ev->button == 1) && (!pd->pager->dragging) &&
(!pd->pager->just_dragged))
2005-09-20 10:09:27 -07:00
{
current_desk = pd->desk;
e_desk_show(pd->desk);
pd->drag.start = 0;
pd->drag.in_pager = 0;
}
pd->pager->just_dragged = 0;
2005-09-20 10:09:27 -07:00
if ((p->popup) && (p->popup->urgent)) _pager_popup_free(p->popup);
}
static void
_pager_desk_cb_mouse_move(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info)
{
Evas_Event_Mouse_Move *ev;
Pager_Desk *pd;
Evas_Coord dx, dy;
unsigned int resist = 0;
E_Drag *drag;
Evas_Object *o, *oo, *o_icon;
Evas_Coord x, y, w, h;
const char *drag_types[] = { "enlightenment/vdesktop" };
Pager_Win *pw;
Eina_List *l;
ev = event_info;
pd = data;
if (!pd) return;
/* prevent drag for a few pixels */
if (pd->drag.start)
{
dx = pd->drag.x - ev->cur.output.x;
dy = pd->drag.y - ev->cur.output.y;
if ((pd->pager) && (pd->pager->inst))
resist = pager_config->drag_resist;
if (((unsigned int) (dx * dx) + (unsigned int) (dy * dy)) <=
(resist * resist)) return;
2010-12-22 15:23:47 -08:00
if (pd->pager) pd->pager->dragging = 1;
pd->drag.start = 0;
}
if (pd->drag.in_pager && pd->pager)
{
evas_object_geometry_get(pd->o_desk, &x, &y, &w, &h);
drag = e_drag_new(pd->pager->zone->container,
x, y, drag_types, 1, pd, -1,
NULL, _pager_desk_cb_drag_finished);
/* set a background to the drag icon */
o = evas_object_rectangle_add(drag->evas);
evas_object_color_set(o, 255, 255, 255, 255);
evas_object_resize(o, w, h);
evas_object_show(o);
/* redraw the desktop theme above */
o = edje_object_add(drag->evas);
e_theme_edje_object_set(o, "base/theme/modules/pager",
"e/modules/pager/desk");
evas_object_show(o);
e_drag_object_set(drag, o);
/* and redraw is content */
oo = e_layout_add(drag->evas);
e_layout_virtual_size_set(oo, pd->pager->zone->w, pd->pager->zone->h);
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
edje_object_part_swallow(o, "e.swallow.content", oo);
evas_object_show(oo);
EINA_LIST_FOREACH(pd->wins, l, pw)
{
int zx, zy;
if ((!pw) || (pw->border->iconic)
|| (pw->border->client.netwm.state.skip_pager))
continue;
o = edje_object_add(drag->evas);
e_theme_edje_object_set(o, "base/theme/modules/pager",
"e/modules/pager/window");
e_layout_pack(oo, o);
e_layout_child_raise(o);
e_zone_useful_geometry_get(pw->desk->desk->zone,
&zx, &zy, NULL, NULL);
e_layout_child_move(o, pw->border->x - zx, pw->border->y - zy);
e_layout_child_resize(o, pw->border->w, pw->border->h);
evas_object_show(o);
if ((o_icon = e_border_icon_add(pw->border, drag->evas)))
{
evas_object_show(o_icon);
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
edje_object_part_swallow(o, "e.swallow.icon", o_icon);
}
}
e_drag_resize(drag, w, h);
e_drag_start(drag, x - pd->drag.dx, y - pd->drag.dy);
pd->drag.from_pager = pd->pager;
pd->drag.from_pager->dragging = 1;
pd->drag.in_pager = 0;
}
}
static void
_pager_desk_cb_drag_finished(E_Drag *drag, int dropped)
{
Pager_Desk *pd;
Pager_Desk *pd2 = NULL;
Eina_List *l;
E_Desk *desk;
E_Zone *zone;
Pager *p;
pd = drag->data;
if (!pd) return;
if (!dropped)
{
/* wasn't dropped on pager, switch with current desktop */
if (!pd->desk) return;
zone = e_util_zone_current_get(e_manager_current_get());
desk = e_desk_current_get(zone);
for (l = pagers; l && !pd2; l = l->next)
{
p = l->data;
pd2 = _pager_desk_find(p, desk);
}
_pager_desk_switch(pd, pd2);
}
if (pd->drag.from_pager)
{
pd->drag.from_pager->dragging = 0;
pd->drag.from_pager->just_dragged = 0;
}
pd->drag.from_pager = NULL;
if (act_popup)
{
e_grabinput_get(input_window, 0, input_window);
if (!hold_count) _pager_popup_hide(1);
}
2005-09-20 10:09:27 -07:00
}
static void
_pager_desk_cb_mouse_wheel(void *data, Evas *e __UNUSED__, Evas_Object *obj __UNUSED__, void *event_info)
2005-09-20 10:09:27 -07:00
{
Evas_Event_Mouse_Wheel *ev;
Pager_Desk *pd;
2005-09-20 10:09:27 -07:00
ev = event_info;
pd = data;
if (pd->pager->popup) return;
if (pager_config->flip_desk)
e_zone_desk_linear_flip_by(pd->desk->zone, ev->z);
}
static Eina_Bool
_pager_popup_cb_timeout(void *data)
{
Pager_Popup *pp;
pp = data;
pp->timer = NULL;
_pager_popup_free(pp);
if (input_window)
{
ecore_x_window_free(input_window);
e_grabinput_release(input_window, input_window);
input_window = 0;
}
return ECORE_CALLBACK_CANCEL;
}
2005-09-20 10:09:27 -07:00
/************************************************************************/
/* popup-on-keyaction functions */
static int
_pager_popup_show(void)
{
E_Zone *zone;
int x, y, w, h;
Pager_Popup *pp;
const char *drop[] =
{
"enlightenment/pager_win", "enlightenment/border",
"enlightenment/vdesktop"
};
if ((act_popup) || (input_window)) return 0;
zone = e_util_zone_current_get(e_manager_current_get());
pp = _pager_popup_find(zone);
if (pp) _pager_popup_free(pp);
input_window = ecore_x_window_input_new(zone->container->win, 0, 0, 1, 1);
ecore_x_window_show(input_window);
if (!e_grabinput_get(input_window, 0, input_window))
{
ecore_x_window_free(input_window);
input_window = 0;
return 0;
}
handlers = eina_list_append
(handlers, ecore_event_handler_add
(ECORE_EVENT_KEY_DOWN, _pager_popup_cb_key_down, NULL));
handlers = eina_list_append
(handlers, ecore_event_handler_add
(ECORE_EVENT_KEY_UP, _pager_popup_cb_key_up, NULL));
handlers = eina_list_append
(handlers, ecore_event_handler_add
(ECORE_EVENT_MOUSE_BUTTON_DOWN, _pager_popup_cb_mouse_down, NULL));
handlers = eina_list_append
(handlers, ecore_event_handler_add
(ECORE_EVENT_MOUSE_BUTTON_UP, _pager_popup_cb_mouse_up, NULL));
handlers = eina_list_append
(handlers, ecore_event_handler_add
(ECORE_EVENT_MOUSE_WHEEL, _pager_popup_cb_mouse_wheel, NULL));
handlers = eina_list_append
(handlers, ecore_event_handler_add
(ECORE_EVENT_MOUSE_MOVE, _pager_popup_cb_mouse_move, NULL));
act_popup = _pager_popup_new(zone, 1);
e_popup_ignore_events_set(act_popup->popup, 1);
evas_object_geometry_get(act_popup->pager->o_table, &x, &y, &w, &h);
e_drop_handler_add(E_OBJECT(act_popup->popup), act_popup->pager,
_pager_drop_cb_enter, _pager_drop_cb_move,
_pager_drop_cb_leave, _pager_drop_cb_drop,
drop, 3, x, y, w, h);
e_bindings_key_grab(E_BINDING_CONTEXT_POPUP, act_popup->popup->evas_win);
evas_event_freeze(act_popup->popup->evas);
evas_event_feed_mouse_in(act_popup->popup->evas,
ecore_x_current_time_get(), NULL);
evas_event_feed_mouse_move(act_popup->popup->evas, -1000000, -1000000,
ecore_x_current_time_get(), NULL);
evas_event_thaw(act_popup->popup->evas);
current_desk = e_desk_current_get(zone);
return 1;
}
static void
_pager_popup_hide(int switch_desk)
{
e_bindings_key_ungrab(E_BINDING_CONTEXT_POPUP, act_popup->popup->evas_win);
hold_count = 0;
hold_mod = 0;
while (handlers)
{
ecore_event_handler_del(handlers->data);
handlers = eina_list_remove_list(handlers, handlers);
}
act_popup->timer = ecore_timer_add(0.1, _pager_popup_cb_timeout, act_popup);
if ((switch_desk) && (current_desk)) e_desk_show(current_desk);
act_popup = NULL;
}
static void
_pager_popup_modifiers_set(int mod)
{
if (!act_popup) return;
hold_mod = mod;
hold_count = 0;
if (hold_mod & ECORE_EVENT_MODIFIER_SHIFT) hold_count++;
if (hold_mod & ECORE_EVENT_MODIFIER_CTRL) hold_count++;
if (hold_mod & ECORE_EVENT_MODIFIER_ALT) hold_count++;
if (hold_mod & ECORE_EVENT_MODIFIER_WIN) hold_count++;
}
static void
_pager_popup_desk_switch(int x, int y)
{
int max_x,max_y, desk_x, desk_y;
Pager_Desk *pd;
Pager_Popup *pp = act_popup;
e_zone_desk_count_get(pp->pager->zone, &max_x, &max_y);
desk_x = current_desk->x + x;
desk_y = current_desk->y + y;
if (desk_x < 0)
desk_x = max_x - 1;
else if (desk_x >= max_x)
desk_x = 0;
if (desk_y < 0)
desk_y = max_y - 1;
else if (desk_y >= max_y)
desk_y = 0;
current_desk = e_desk_at_xy_get(pp->pager->zone, desk_x, desk_y);
pd = _pager_desk_find(pp->pager, current_desk);
if (pd) _pager_desk_select(pd);
edje_object_part_text_set(pp->o_bg, "e.text.label", current_desk->name);
}
static void
_pager_popup_cb_action_show(E_Object *obj __UNUSED__, const char *params __UNUSED__, Ecore_Event_Key *ev __UNUSED__)
{
if (_pager_popup_show())
_pager_popup_modifiers_set(ev->modifiers);
}
static void
_pager_popup_cb_action_switch(E_Object *obj __UNUSED__, const char *params, Ecore_Event_Key *ev)
{
int max_x,max_y, desk_x;
int x = 0, y = 0;
if (!act_popup)
{
if (_pager_popup_show())
_pager_popup_modifiers_set(ev->modifiers);
else
return;
}
e_zone_desk_count_get(act_popup->pager->zone, &max_x, &max_y);
2010-12-22 14:41:26 -08:00
desk_x = current_desk->x /* + x <=this is always 0 */;
if (!strcmp(params, "left"))
x = -1;
else if (!strcmp(params, "right"))
x = 1;
else if (!strcmp(params, "up"))
y = -1;
else if (!strcmp(params, "down"))
y = 1;
else if (!strcmp(params, "next"))
{
x = 1;
if (desk_x == max_x - 1)
y = 1;
}
else if (!strcmp(params, "prev"))
{
x = -1;
if (desk_x == 0)
y = -1;
}
_pager_popup_desk_switch(x, y);
}
static Eina_Bool
_pager_popup_cb_mouse_down(void *data __UNUSED__, int type __UNUSED__, void *event)
{
Ecore_Event_Mouse_Button *ev;
Pager_Popup *pp = act_popup;
ev = event;
if (ev->window != input_window) return ECORE_CALLBACK_PASS_ON;
evas_event_feed_mouse_down(pp->popup->evas, ev->buttons,
0, ev->timestamp, NULL);
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_popup_cb_mouse_up(void *data __UNUSED__, int type __UNUSED__, void *event)
{
Ecore_Event_Mouse_Button *ev;
Pager_Popup *pp = act_popup;
ev = event;
if (ev->window != input_window) return ECORE_CALLBACK_PASS_ON;
evas_event_feed_mouse_up(pp->popup->evas, ev->buttons,
0, ev->timestamp, NULL);
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_popup_cb_mouse_move(void *data __UNUSED__, int type __UNUSED__, void *event)
{
Ecore_Event_Mouse_Move *ev;
Pager_Popup *pp = act_popup;
ev = event;
if (ev->window != input_window) return 1;
evas_event_feed_mouse_move(pp->popup->evas,
ev->x - pp->popup->x + pp->pager->zone->x,
ev->y - pp->popup->y + pp->pager->zone->y,
ev->timestamp, NULL);
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_popup_cb_mouse_wheel(void *data __UNUSED__, int type __UNUSED__, void *event)
{
Ecore_Event_Mouse_Wheel *ev = event;
Pager_Popup *pp = act_popup;
int max_x;
e_zone_desk_count_get(pp->pager->zone, &max_x, NULL);
if (current_desk->x + ev->z >= max_x)
2008-01-07 16:45:12 -08:00
_pager_popup_desk_switch(1, 1);
else if (current_desk->x + ev->z < 0)
2008-01-07 16:45:12 -08:00
_pager_popup_desk_switch(-1, -1);
else
2008-01-07 16:45:12 -08:00
_pager_popup_desk_switch(ev->z, 0);
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_popup_cb_key_down(void *data __UNUSED__, int type __UNUSED__, void *event)
{
Ecore_Event_Key *ev;
ev = event;
if (ev->window != input_window) return ECORE_CALLBACK_PASS_ON;
if (!strcmp(ev->key, "Up"))
_pager_popup_desk_switch(0, -1);
else if (!strcmp(ev->key, "Down"))
_pager_popup_desk_switch(0, 1);
else if (!strcmp(ev->key, "Left"))
_pager_popup_desk_switch(-1, 0);
else if (!strcmp(ev->key, "Right"))
_pager_popup_desk_switch(1, 0);
else if (!strcmp(ev->key, "Escape"))
_pager_popup_hide(0);
else
{
E_Config_Binding_Key *binding;
Eina_List *l;
EINA_LIST_FOREACH(e_config->key_bindings, l, binding)
{
E_Binding_Modifier mod = 0;
if ((binding->action) && (strcmp(binding->action,"pager_switch")))
continue;
if (ev->modifiers & ECORE_EVENT_MODIFIER_SHIFT)
mod |= E_BINDING_MODIFIER_SHIFT;
if (ev->modifiers & ECORE_EVENT_MODIFIER_CTRL)
mod |= E_BINDING_MODIFIER_CTRL;
if (ev->modifiers & ECORE_EVENT_MODIFIER_ALT)
mod |= E_BINDING_MODIFIER_ALT;
if (ev->modifiers & ECORE_EVENT_MODIFIER_WIN)
mod |= E_BINDING_MODIFIER_WIN;
if (binding->key && (!strcmp(binding->key, ev->keyname)) &&
((binding->modifiers == mod)))
{
E_Action *act;
act = e_action_find(binding->action);
if (act)
{
if (act->func.go_key)
act->func.go_key(NULL, binding->params, ev);
}
}
}
}
return ECORE_CALLBACK_PASS_ON;
}
static Eina_Bool
_pager_popup_cb_key_up(void *data __UNUSED__, int type __UNUSED__, void *event)
{
Ecore_Event_Key *ev;
ev = event;
if (!(act_popup)) return ECORE_CALLBACK_PASS_ON;
if (hold_mod)
{
if ((hold_mod & ECORE_EVENT_MODIFIER_SHIFT) &&
(!strcmp(ev->key, "Shift_L"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_SHIFT) &&
(!strcmp(ev->key, "Shift_R"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_CTRL) &&
(!strcmp(ev->key, "Control_L"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_CTRL) &&
(!strcmp(ev->key, "Control_R"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_ALT) &&
(!strcmp(ev->key, "Alt_L"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_ALT) &&
(!strcmp(ev->key, "Alt_R"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_ALT) &&
(!strcmp(ev->key, "Meta_L"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_ALT) &&
(!strcmp(ev->key, "Meta_R"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_ALT) &&
(!strcmp(ev->key, "Super_L"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_ALT) &&
(!strcmp(ev->key, "Super_R"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_WIN) &&
(!strcmp(ev->key, "Super_L"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_WIN) &&
(!strcmp(ev->key, "Super_R"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_WIN) &&
(!strcmp(ev->key, "Mode_switch"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_WIN) &&
(!strcmp(ev->key, "Meta_L"))) hold_count--;
else if ((hold_mod & ECORE_EVENT_MODIFIER_WIN) &&
(!strcmp(ev->key, "Meta_R"))) hold_count--;
if ((hold_count <= 0) && (!act_popup->pager->dragging))
{
_pager_popup_hide(1);
return ECORE_CALLBACK_PASS_ON;
}
}
return ECORE_CALLBACK_PASS_ON;
}
/***************************************************************************/
/* module setup */
EAPI E_Module_Api e_modapi =
{
E_MODULE_API_VERSION, "Pager"
};
2005-09-20 10:09:27 -07:00
EAPI void *
e_modapi_init(E_Module *m)
{
conf_edd = E_CONFIG_DD_NEW("Pager_Config", Config);
#undef T
#undef D
#define T Config
#define D conf_edd
E_CONFIG_VAL(D, T, popup, UINT);
E_CONFIG_VAL(D, T, popup_speed, DOUBLE);
E_CONFIG_VAL(D, T, popup_urgent, UINT);
E_CONFIG_VAL(D, T, popup_urgent_stick, UINT);
E_CONFIG_VAL(D, T, popup_urgent_speed, DOUBLE);
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
E_CONFIG_VAL(D, T, show_desk_names, UINT);
E_CONFIG_VAL(D, T, popup_height, INT);
E_CONFIG_VAL(D, T, popup_act_height, INT);
E_CONFIG_VAL(D, T, drag_resist, UINT);
E_CONFIG_VAL(D, T, btn_drag, UCHAR);
E_CONFIG_VAL(D, T, btn_noplace, UCHAR);
E_CONFIG_VAL(D, T, btn_desk, UCHAR);
E_CONFIG_VAL(D, T, flip_desk, UCHAR);
2005-09-20 10:09:27 -07:00
pager_config = e_config_domain_load("module.pager", conf_edd);
2005-09-20 10:09:27 -07:00
if (!pager_config)
{
pager_config = E_NEW(Config, 1);
pager_config->popup = 1;
pager_config->popup_speed = 1.0;
pager_config->popup_urgent = 0;
pager_config->popup_urgent_stick = 0;
pager_config->popup_urgent_speed = 1.5;
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
pager_config->show_desk_names = 1;
pager_config->popup_height = 60;
pager_config->popup_act_height = 60;
pager_config->drag_resist = 3;
pager_config->btn_drag = 1;
pager_config->btn_noplace = 2;
pager_config->btn_desk = 0;
pager_config->flip_desk = 0;
2005-09-20 10:09:27 -07:00
}
E_CONFIG_LIMIT(pager_config->popup, 0, 1);
E_CONFIG_LIMIT(pager_config->popup_speed, 0.1, 10.0);
E_CONFIG_LIMIT(pager_config->popup_urgent, 0, 1);
E_CONFIG_LIMIT(pager_config->popup_urgent_stick, 0, 1);
E_CONFIG_LIMIT(pager_config->popup_urgent_speed, 0.1, 10.0);
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
E_CONFIG_LIMIT(pager_config->show_desk_names, 0, 1);
E_CONFIG_LIMIT(pager_config->popup_height, 20, 200);
E_CONFIG_LIMIT(pager_config->popup_act_height, 20, 200);
E_CONFIG_LIMIT(pager_config->drag_resist, 0, 50);
E_CONFIG_LIMIT(pager_config->flip_desk, 0, 1);
E_CONFIG_LIMIT(pager_config->btn_drag, 0, 32);
E_CONFIG_LIMIT(pager_config->btn_noplace, 0, 32);
E_CONFIG_LIMIT(pager_config->btn_desk, 0, 32);
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_RESIZE, _pager_cb_event_border_resize, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_MOVE, _pager_cb_event_border_move, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_ADD, _pager_cb_event_border_add, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_REMOVE, _pager_cb_event_border_remove, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_ICONIFY, _pager_cb_event_border_iconify, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_UNICONIFY, _pager_cb_event_border_uniconify, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_STICK, _pager_cb_event_border_stick, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_UNSTICK, _pager_cb_event_border_unstick, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_DESK_SET, _pager_cb_event_border_desk_set, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_STACK, _pager_cb_event_border_stack, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_ICON_CHANGE, _pager_cb_event_border_icon_change, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_URGENT_CHANGE, _pager_cb_event_border_urgent_change, NULL));
pager_config->handlers = eina_list_append
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_FOCUS_IN,
_pager_cb_event_border_focus_in, NULL));
pager_config->handlers = eina_list_append
e: 1. configure/build changes to allow cross-compiling painlessly 2. pager module namespace changes - this was still dirty afdter the namespace cleanup, so clean it up 3. add a powersave subsystem - doesnt have an "automatic" way to turn on and off right now, this i think is best provided by modules (that do things like monitor acpi status's (eg close lid of laptop), AC power status etc. etc. this allows e to nicely defer "power" expensive actions to avoid disk spinups etc. 4. move to use the new ecore poller system - discussed long ago as part of power management/saving issues. now it exists 5. add a canvas idle flush call that helsp cope with the new shm greedy software x11 engine stuff 6. use the new powersave subsystem where appropriate 7. fix non-zeroed/initted memory access in e_fm_main 8. fix mem leak for e menus 9. remove ipc handlers for changed/removed config values 10. use animaotr not timer for menu scrolls - then menu scrolls obey the fps config 11. fix up timer/poll happienss of cursor idle stuff 12. remove avoid damage from popups for now - causing problems 13. change battery and temp readouts to b e shorter so they fit 14. pager can emit signals on focus change for mini-windows now 15. temperature module now uses a slave process and uses stdin/out to talk to it and get output - this makes e smoother as in my expereicne i found getting the temp on my laptop actually took like 200ms so e "hang" for 200ms while reading the acpi files - so now the subprocess does it and just writesa back to e when it gets it. ecore: 1. add ecore_pollers. see the documentation on them in doxygen comments :) 2. fix timers to only go off when they have to - bug there that made e's select time out a LOT more than it needed to. defensive coding hid the problem. now fixed. e should be much more power friendly now. 3. formatting/niceness in ecore_exe stuff 4. some comments on comments with SIGIO ideas vs. select 5. add call to be able to add an idle enterer at the start of the list of them, not just the end (as has been the default) 6. fix ecore_evas to support auto evas idler calls after 0.5 secs of idle in all canvases - and to do it right 7. if argb destination - set the shape EVENT shape (to mask out events in transparent regions much like shape does withotu translucency) 8. in ecore_x add support for the event shape evas: 1. fix cache to work properly and not just always fill up (as it seemed to like to think cahce useage dropped below 0 when it didnt and thus just over-fill) 2. software x11 engine now ONLY uses shm segments - no ximages over the socket. this ximage hack was there to avoid the 2 round trips involved in setting up an shm image - now i mitigated that wih an shm image cache pool. it keeps shm images around and repurposes them for new update regions if appropriate. this means many fewer shm creates (about 1/100th the number) and since we recycle the memory less 0 memory page filling by the kernel - in the end, i recorded about a 10-20% speedup over the old software x11 engine. simple tests i have seen up to 120% speedups. idle flush now does something - it frees all the cached shm segments. it has a hard-coded limit of 4mb worth of shm segments (or 32 segments - whichever comes first) to keep around. once can never complain much about speedups methinks :). also evas will defer sync until the NEXT frame is written - this means evas can calculate the next frame of data while x dma's/copies the images to the screen at the same time (if you hve a dual core or multi-cpu machnike or your xserver is able to use DMA to copy image data to the screen/video ram then this should see a decent speedup). SVN revision: 33448
2008-01-10 23:33:57 -08:00
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_FOCUS_OUT,
_pager_cb_event_border_focus_out, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_BORDER_PROPERTY, _pager_cb_event_border_property, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_ZONE_DESK_COUNT_SET, _pager_cb_event_zone_desk_count_set, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_DESK_SHOW, _pager_cb_event_desk_show, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_DESK_NAME_CHANGE, _pager_cb_event_desk_name_change, NULL));
pager_config->handlers = eina_list_append
(pager_config->handlers, ecore_event_handler_add
(E_EVENT_CONTAINER_RESIZE, _pager_cb_event_container_resize, NULL));
pager_config->module = m;
2006-05-31 09:19:11 -07:00
e_gadcon_provider_register(&_gadcon_class);
e_configure_registry_item_add("extensions/pager", 40, N_("Pager"), NULL,
"preferences-pager", _pager_config_dialog);
act_popup_show = e_action_add("pager_show");
if (act_popup_show)
{
act_popup_show->func.go_key = _pager_popup_cb_action_show;
e_action_predef_name_set(N_("Pager"), N_("Show Pager Popup"),
"pager_show", "<none>", NULL, 0);
}
act_popup_switch = e_action_add("pager_switch");
if (act_popup_switch)
{
act_popup_switch->func.go_key = _pager_popup_cb_action_switch;
e_action_predef_name_set(N_("Pager"), N_("Popup Desk Right"),
"pager_switch", "right", NULL, 0);
e_action_predef_name_set(N_("Pager"), N_("Popup Desk Left"),
"pager_switch", "left", NULL, 0);
e_action_predef_name_set(N_("Pager"), N_("Popup Desk Up"),
"pager_switch", "up", NULL, 0);
e_action_predef_name_set(N_("Pager"), N_("Popup Desk Down"),
"pager_switch", "down", NULL, 0);
e_action_predef_name_set(N_("Pager"), N_("Popup Desk Next"),
"pager_switch", "next", NULL, 0);
e_action_predef_name_set(N_("Pager"), N_("Popup Desk Previous"),
"pager_switch", "prev", NULL, 0);
}
return m;
2005-09-20 10:09:27 -07:00
}
EAPI int
e_modapi_shutdown(E_Module *m __UNUSED__)
{
e_gadcon_provider_unregister(&_gadcon_class);
if (pager_config->config_dialog)
e_object_del(E_OBJECT(pager_config->config_dialog));
while (pager_config->handlers)
{
ecore_event_handler_del(pager_config->handlers->data);
pager_config->handlers =
eina_list_remove_list(pager_config->handlers, pager_config->handlers);
}
if (pager_config->menu)
{
e_menu_post_deactivate_callback_set(pager_config->menu, NULL, NULL);
e_object_del(E_OBJECT(pager_config->menu));
pager_config->menu = NULL;
}
e_configure_registry_item_del("extensions/pager");
e_action_del("pager_show");
e_action_del("pager_switch");
e_action_predef_name_del(N_("Pager"), N_("Popup Desk Right"));
e_action_predef_name_del(N_("Pager"), N_("Popup Desk Left"));
e_action_predef_name_del(N_("Pager"), N_("Popup Desk Up"));
e_action_predef_name_del(N_("Pager"), N_("Popup Desk Down"));
e_action_predef_name_del(N_("Pager"), N_("Popup Desk Next"));
e_action_predef_name_del(N_("Pager"), N_("Popup Desk Previous"));
E_FREE(pager_config);
E_CONFIG_DD_FREE(conf_edd);
return 1;
}
EAPI int
e_modapi_save(E_Module *m __UNUSED__)
{
e_config_domain_save("module.pager", conf_edd, pager_config);
return 1;
}